There is momentum building behind the adoption of pre-print servers in the life sciences. Ron Vale, a professor of cellular and molecular pharmacology at UCSF and Lasker Award winner, has just added a further powerful impulse to this movement in the form, appropriately, of a pre-print posted to the bioRxiv just a few days ago.
If you are a researcher and haven’t yet thought seriously about pre-prints, please read Vale’s article. It is thoughtful and accessible (and there is a funny section in which he imagines the response of modern-day reviewers to Watson and Crick’s 1953 paper on the structure of DNA). His reasoning is built on the concern that there has been a perceptible increase over the last thirty years in the amount of experimental data – and therefore work – required for PhD students to get their first major publication. Vale argues that this is a result of the increased competition within the life sciences, which is focused on restricted access to ‘top journals’ and is in turn due to the powerful hold that journal impact factors now have over people’s careers. Regulars readers will know that the problems with impact factors are a familiar topic on this blog (and may even be aware that their mis-use was one of the issues highlighted in The Metric Tide, the report of the HEFCE review of the use of metrics in research assessment that was published last week).
Michael Eisen has written a sympathetic critique of Vale’s paper. He takes some issue with the particulars of the arguments about increased data requirements but nevertheless espouses strong support for the drive – which he has long pursued himself – for more rapid forms of publication.
I won’t bother to rehearse Eisen’s critique since I think it is the bigger picture that warrants most attention. This bigger picture – the harmful effect of the chase after impact factors on the vitality and efficiency of scientific community – emerged as a central theme at the Royal Society meeting convened earlier this year to discuss the Future of Scholarly Scientific Communication. At that gathering I detected a palpable sense among attendees that the wider adoption of pre-print servers would be an effective and feasible way to improve the dissemination of research results; (for more background, see proposal 3 towards the bottom of my digest of the meeting).
Vale’s article does an excellent job of articulating the support for pre-prints that bloomed at the Royal Society meeting. I urge you again to read it. But for the tl;dr (“too long; didn’t read”) crowd, here’s the key section on the arguments for pre-prints (with my emphases in boldface)*:
1) Submission to a pre-print repository would allow a paper to be seen and evaluated by colleagues and search/grant committees immediately after its completion. This could enable trainees to apply for postdoctoral positions, grants, or jobs earlier than waiting for the final journal publication. A recent study of several journals found an average delay of ~7 months from acceptance to publication (33), but this is average depended upon the journal and the review/revision process can take longer on a case-by-case basis. Furthermore, this time does not take rejections into account and the potential need to “shop” for a journal that will publish the work.
2) A primary objective of a pre-print repository is to transmit scientific results more rapidly to the scientific community, which should appeal to funding agencies whose main objective is to catalyze new discoveries overall. Furthermore, authors receive faster and broader feedback on their work than occurs through peer review, which can help advance their own studies.
3) If widely adopted, a pre-print repository (which acts an umbrella to collect all scientific work and is not associated** with any specific journal) could have the welcoming effect of having colleagues read and evaluate scientific work well before it has been branded with a journal name. Physicists tend to rely less on journal impact factors for evaluation, in part, because they are used to reading and evaluating science posted on arXiv. Indeed, some major breakthroughs posted on arXiv were never published subsequently in a journal. The life science community needs to return to a culture of evaluating scientific merit from reading manuscripts, rather than basing judgment on where papers were published and hence outsourcing the career evaluation process to journals.
4) A pre-print repository may not solve the “amount of data” required for the next step of journal publication. However, it might lower the bar for shorter manuscripts to be posted and reach the community, even if an ensuing submission to a journal takes longer to develop.
5) A pre-print repository is good value in terms of impact and information transferred per dollar spent. Compared to operating a journal, the cost of running arXiv is low (~$800,000 per year), most of which comes from modest subscription payments from 175 institutions and a matching grant from the Simons Foundation. Unlike a journal, submissions to arXiv are free.
6) Future innovations and experiments in peer-to-peer communication and evaluation could be built around an open pre-print server. Indeed, such communications might provide additional information and thus aid journal-based peer review.
7) A pre-print server for biology represents a feasible action item, since the physicists/mathematicians have proof-of-principle that this system works and can co-exist with journals.
The last point is perhaps the most important. Publishing pre-prints is a feasible step. I have started to do it myself in the past year (partly motivated by deals offered by PeerJ) and it is a practice that I intend to continue.
But the key will be to get more and more life scientists to adopt the pre-print habit. Leadership on this from senior figures – academicians, fellows of the Royal Society, prize winners and the like – will help. Institutional support from funders and universities, by which I meant putting in place incentives for rapid communication of results, could also be important. The rest of us have to face this idea seriously and at the very least be willing to debate the pros and cons – I would welcome that discussion.
Or you could see the sense publishing preprints and just do it.
*Thanks to Ron Vale for permission to reproduce this section of his paper.
**In Vale’s article this phrase is ’not unassociated’ but I suspect that’s a typo.
I do completely agree with these thoughts. Not because I am a Physicist and have used the arXiv since more than 20 years. And not because I want to harm publishing industry after having worked more than a decade there. But to enable researchers of all academic disciplines to have immediate access to the results of current research of their peers. This is what’s needed to support scientific discourse and not the question whether I have published my recent work in a fancy, high-IF journal. If those who still support the latter (totally non-scientific) behaviour will understand this message, we have succeeded to establish scholarly communication in the 21st century, And I am afraid that not established publishers are setting up the hurdle but still many researchers themselves. Let’s continue to convince them!
And don’t continue to call it a “preprint” if a new work has been posted in a repository or on a server – it’s research that has been made public – so it’s published literally.
Thanks for the comment Alexander. I suspect you are right that it is researchers themselves who are the biggest barrier to change. We are a conservative bunch and I myself have not considered preprints seriously as a publishing avenue until relatively recently. But they make a great deal of sense and I was struck by the breadth and depth of support for the idea at the Royal Society meeting. Having the likes of Ron Vale wade into the debate can only be a good thing.
Subfields that intersect with biology have already been doing this for a while now – in bioinformatics, it’s very common to see the papers published as preprints at the same time as they are submitted for review. I’d guess 50% of papers that I review are things I’ve already seen on preprint servers.
I should have acknowledged in my piece that there are several branches of the life sciences (and especially bioinformatics) where use of the arXiv is relatively common.
That said, it is also true that not all branches of physics make use of the arXiv which has often struck me as odd. My guess is that the use of preprints is more embedded in those fields where researchers are reliant on access to large-scale international facilities for their work so that collaboration is more natural (and competition less intense?). But if anyone has insights on the varying practices between different sub-disciplines, I’d be interested to see them.
As said earlier I have been working actively working as researcher in different fields of physics for about 10 years before I joined publishing industry. We were small research groups and were used to access arXiv on a daily basis (I frequently confirmed this observation when visiting my colleagues during the recent years). My oberservation it is not the fact that research is done on large scale which may result in a more frequent usage of arXiv. Most research (not only in physics) is done in competing groups worldwide and it appears to be straightforward to find out as soon as possible what your peers are going to publish. Couldn’t it rather be the lack of usage or knowledge about these repositories? The arXiv is a terrific source for an immediate exchange of most recent research information and I wish that more scientists will understand the grat potential of this concept. It is amazing that the BioRxiv was deployed twenty years after the first paper had been posted at the arXiv server.
Many biologists cite their concern of getting scooped as the premier reason not to post on a preprint server – the same reason why many biology conferences are abundant with talks on published or in press data (a big change from years back when conferences were about sharing unpublished data). In essence, preprints are/should be seen as an extension of conferences.
In physics there seems to be a culture of staking a claim on findings via a preprint, while in biology the fear prevails that freeloaders will encroach – maybe because often experiments are very easy/fast to replicate.
Alexander, I think it is crucial to refer to preprint posting as posts and peer review papers as publications – we need to clearly distinguish between these different levels of validation. Peer-review may have faults, but it has served us well and will continue to do so – given the information glut, we need to increasingly rely on filtering and validation processes.
I agree that preprints would serve the biology community very well for all the reasons discussed – it should smoothen the path to publication by moving revision upstream of journal peer review, as is the case in physics, where papers tend to be published much more efficiently (less delay, fewer rounds of revision, less frustration).
I have also heard the concern about being scooped. But the advantage of a pre-print over a conference presentation (talk or poster) is that it is a date-stamped citable object. For this reason, physicists have developed habit of treating pre-print publication as a claim of priority, so is it not odd for life scientists to disregard this possibility? Of course this brings us back to the notion that we are transfixed by the notion that only publications that matter are those in high-IF peer-reviewed journals.
So I don’t think Bernd’s suggestion that pre-print should be referred to as posts and papers as publications is a helpful one. We have got to get to a point where we can judge work on its intrinsic merits and not on the wrapper. Yes, that is easier said than done but, as Vale argues, wider adoption of pre-prints could help us to shift focus.
Bernd’s point is very interesting. Do we actually know what are the main reservations biologists have about preprinting? Has anyone surveyed this? Or maybe looked at fields (bioinformatics/ecology) that have rapidly adopted preprints to figure out why they jumped on board?
From talking to a few people, in addition to being scooped, I think the main concerns are:
1. whether it prevents submitting to a “top journal” – it doesn’t.
2. not wanting to put something up on the web that might be wrong
3. not wanting to put something up on the web that isn’t formatted nicely like a paper(!) – there’s even an template from Dennis Eckmeier to get a pro-look to your preprint.
4. plain old conservatism – which is odd… aren’t we supposed to be breaking the mould?
With regard to point 1, this is a declining problem. There is some guidance on the bioRxiv site:
Many research journals, including all Cold Spring Harbor Laboratory Press titles (Genome Research, Genes & Development, and Learning & Memory), EMBO Journal, Nature journals, Science, eLife, and all PLOS journals allow posting on preprint servers such as bioRxiv prior to publication. A few journals will not consider articles that have been posted to preprint servers. A list of journal policies can be found on Wikipedia and SHERPA/RoMEO. Authors should consult these lists and other sources of information before posting on bioRxiv.
I wonder how many is ‘many’ – and whether it would be possible to compile a list of non-compliant journals and then whittle away at it?
In my own field the exception is J Cell Sci which has still not explicitly said that preprints are OK (although other CoB journals have changed their policy). J Cell Biol changing their policy is very significant for us cell biologists. I agree that identifying non-compliant journals and whittling away is a good strategy. I’ve talked to J Cell Sci about it in the past and they claimed there was no demand to change their policy, because I was the only person who had asked!
Apologies, I should have posted this comment and my previous under my own name (Steve Royle) and declared that I am an Affiliate for bioRxiv.
At this point I should probably also declare that I am an Affiliate for bioRxiv…
Stephen, I agree that preprints serve an important purpose in reducing the lag in communicating science and in improving it based on comments from the community. The citable time stamp to stake a claim is indeed a reason cited for the success in physics.
quantixed is right that many still worry that a preprint post prevent publication in certain journals, but I am not aware of any reputable journals in biology that would reject a manuscript on this basis – since some of the policies only changed recently this will take some time to sink in.
Journals in fact adopted these policies based on the premise that peer review adds value and I would certainly agree with that – this is why it is important to differentiate between peer reviewed and non-peer reviewed archival content. This is certainly not to say that the latter should not inform research progress and indeed that it should not also count in research assessment. As you say, it is the content that matters – but with >1.5 million papers/year we do need reliable filtering mechanisms.
Maybe at some stage community commenting at preprint level will fulfil that role, but that will take a little longer. First folks should start experimenting with posting their work on preprints, as you have done.
I would like to go even further by saying that the notion of “print” is not one inherently associated with Science but with journals and copyright transfers. What turns a random article into a scientific article is the consensus of the scientific community, not the decision of one guy sitting in an editorial board, taken through an unverifiable process. Therefore, I would like to concur with Ron Vale by saying that sharing our “pre-prints” (i.e what we think) and having them discussed and debated openly in a scientific public place is the only way way to do science properly. Having them stamped by one impact-factor journal is what you need to get a grant, and definitely not what you need to do science. Because it is scientific, a scientific article must always be open to debate. It never has a final version (the “print”) but only a current one. It must be able to evolve as a consensus gradually forms around it. Thanks to Internet, this fundamental ethics of Science can now be straightforwardly implemented on pre-print servers, and since it is no longer incompatible with getting a grant (by giving copyrights to journals afterwards), we should definitely do it.
Pre-print servers also allow for the implementation of community-wide curation mechanisms that will offer sooner or later the necessary alternative to impact factor, one based on judgment and not on “traffic data” such as citations.
To prove that I am a man of action and not just words, the latest manuscript from my lab – which was submitted for publication yesterday morning – has also been deposited in the bioRxiv. Those interested in our new findings about the molecular mechanism of translation initiation in noroviruses can read about them without delay.